Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=rnjm20

Nordic Journal of Music Therapy

ISSN: 0809-8131 (Print) 1944-8260 (Online) Journal homepage: https://www.tandfonline.com/loi/rnjm20

Effects of live music therapy on heart ratevariability and self-reported stress and anxietyamong hospitalized pregnant women: Arandomized controlled trial

Pia Teckenberg-Jansson, Siiri Turunen, Tarja Pölkki, Minna-Johanna Lauri-Haikala, Jari Lipsanen, Andreas Henelius, Ansa Aitokallio-Tallberg, SatuPakarinen, Marianne Leinikka & Minna Huotilainen

To cite this article: Pia Teckenberg-Jansson, Siiri Turunen, Tarja Pölkki, Minna-Johanna Lauri-Haikala, Jari Lipsanen, Andreas Henelius, Ansa Aitokallio-Tallberg, Satu Pakarinen, MarianneLeinikka & Minna Huotilainen (2019) Effects of live music therapy on heart rate variability and self-reported stress and anxiety among hospitalized pregnant women: A randomized controlled trial,Nordic Journal of Music Therapy, 28:1, 7-26, DOI: 10.1080/08098131.2018.1546223

To link to this article: https://doi.org/10.1080/08098131.2018.1546223

View supplementary material Published online: 24 Jan 2019.

Submit your article to this journal Article views: 382

View related articles View Crossmark data

Citing articles: 1 View citing articles

ORIGINAL RESEARCH ARTICLE

Effects of live music therapy on heart rate variability andself-reported stress and anxiety among hospitalizedpregnant women: A randomized controlled trialPia Teckenberg-Janssona, Siiri Turunenb, Tarja Pölkkic, Minna-Johanna Lauri-Haikalab, Jari Lipsanenb, Andreas Heneliusd,e, Ansa Aitokallio-Tallbergf,Satu Pakarinend, Marianne Leinikkad and Minna Huotilainenb

aSävelvoima-Tonkraft, Västerskog, Finland; bCICERO Learning Network, Faculty of EducationalSciences and Cognitive Brain Research Unit, Department of Psychology and Logopedics, Faculty ofMedicine, University of Helsinki, Helsinki, Finland; cDepartment of Children and Women, OuluUniversity Hospital, Oulu, Finland; dFinnish Institute of Occupational Health, Helsinki, Finland;eDepartment of Computer Science, University of Helsinki, Helsinki, Finland; fDepartment ofObstetrics and Gynecology, Helsinki University Hospital, University of Helsinki, Helsinki, Finland

ABSTRACTIntroduction: This study aims to evaluate the effects of live music therapy on heart ratevariability (HRV) and self-reported stress and anxiety among hospitalized women withhigh-risk pregnancies. A total of 102 women at an antenatal ward due to pregnancy-related complications participated in a randomized controlled trial.Methods: The participants were randomly assigned to a music therapy group (N = 52) orcontrol group (N=50). Thewomen in themusic therapy group received livemusic therapyon three consecutive days, for half an hour at a time. The participants belonging to thecontrol group were instructed to rest for equally long time periods. The physiologic stressof the participants was assessed using HRV measures. The participants also rated theirperceived stress and anxiety. The physiologic stress of the participants was assessed using12 HRV measures.Results: The SD2measure of HRV increased significantlymore in themusic therapy groupthan in the control group during the therapy sessions. Moreover, the low frequency (LF)HRVmeasure decreased during the three-day therapy period. The self-reported stress wasnot significantly altered after the intervention. For women with high initial self-reportedanxiety in both groups, their anxietywas significantly reducedduring the three-day period.

ARTICLE HISTORY Received 9 September 2017; Accepted 26 October 2018

KEYWORDS Prenatal stress; prenatal anxiety; live music therapy; pregnancy complications; heart ratevariability (HRV)

Introduction

Research over the last three decades has established that women experiencing highstress and anxiety during pregnancy are at risk of preterm birth and miscarriage

CONTACT Pia Teckenberg-Jansson pia.teckenberg@kolumbus.fi Sävelvoima-Tonkraft, Stuvunäsvägen11, 01120 Västerskog, Finland; Minna Huotilainen minna.huotilainen@helsinki.fi CICERO LearningNetwork, Faculty of Educational Sciences, P.O. Box 9, 00014 University of Helsinki, Finland.

Supplementary data for this article can be accessed here.© 2018 GAMUT – The Grieg Academy Music Therapy Research Centre

NORDIC JOURNAL OF MUSIC THERAPY2019, VOL. 28, NO. 1, 7–26https://doi.org/10.1080/08098131.2018.1546223

(Dole et al., 2003; Douglas, 2010; Lobel et al., 2008; Mulder et al., 2002). Preterm bornchildren with low birth weight may suffer from long-term health problems such asneurodevelopmental disorders (Soleimani, Zaheri, & Abdi, 2014) and later depression(Räikkönen et al., 2008). High antenatal maternal stress and anxiety is related toregulation problems at the cognitive, behavioral, and emotional levels. Further, stressand anxiety may interfere with fetal development, even increasing the risk for severelifelong diseases (Douglas, 2010; Van Den Bergh, Mulder, Mennes, & Glover, 2005).Maternal stress has also been shown to have a causal connection to fetal stress asreduced fetal HRV (DiPietro, Hodgson, Costigan, Hilton, & Johnson, 1996).

Pregnancy-specific stress is indicated to be a more powerful contributor to birthoutcomes than general stress (Lobel et al., 2008), and having a high-risk pregnancyincreases the impact of stress (Geller, 2004). Notably, hospitalized pregnant womentreated with bed rest have a considerably higher risk of experiencing anxiety andstress (Maloni & Park, 2005). Physical and emotional changes during pregnancy, inaddition to the concerns about parenting, changing family relationships, labor, andthe child’s health, may produce stress for pregnant women (Chang, Chen, & Huang,2008, Lobel et al., 2008). Prenatal stress and anxiety should be seriously considered inhealth care due to its consequences for fetal development and well-being (Dennis,Falah-Hassani, & Shiri, 2017; Graignic-Philippe & Tordjman, 2009; Van Den Berghet al., 2005). Consequently, there is an essential need for stress relief duringpregnancy.

Non-pharmacological interventions have been developed specifically for the careof pregnant women, since pharmacological treatments may pose serious risks forboth the mother and fetus (Calderon-Margalit, Qiu, Ornoy, Siscovick, & Williams,2009; Chang et al., 2008). Bed rest is often prescribed for women with high-riskpregnancies but may result in adverse physical and psychological consequences suchas fatigue, backache, and depression, which may continue after the child is born(Maloni & Park, 2005). Maternal relaxation may be beneficial for the fetus and lead todecreased fetal heart rate, increased fetal HRV and suppression of fetal motor activity(Dipietro, Costigan, Nelson, Gurewitsch, & Laudenslager, 2008). Significant prolon-gation of pregnancy can be achieved through a relaxation training program forhospitalized pregnant women (Chuang et al., 2012a). Beneficial effects of relaxationon reducing self-reported anxiety and stress in pregnant women have also been found(Bastani, Hidarnia, Kazemnejad, Vafaei, & Kashanian, 2005).

Music is widely known to enhance well-being, reduce stress and distract patientsfrom unpleasant symptoms by exerting direct physiological effects through theautonomic nervous system (Kemper & Danhauer, 2005). Listening to relaxingmusic lowers perceived stress levels, as indicated by reduced cortisol levels, loweredheart rate, and lowered mean arterial pressure (Burrai, Hasan, Fancourt, Luppi, &DiSomma, 2016; Lai & Li, 2011). Bieligmeyer (2015) studied the effects of the Taosound-bed (Benedek) on oncology patients and found significant changes of innerbalance, vitality, and vigilance. Gilboa (2014) studied the uterine auditory environ-ment and its effects on the fetus. Mothers listening to music during pregnancyexperienced lower levels of psychosocial stress; music particularly alleviated the stressrelated to future baby care, changing family relationships, and the maternal roleidentification (Chang, Yu, Chen, & Chen, 2015).

8 P. TECKENBERG-JANSSON ET AL.

Music therapy has previously been used in the care of pregnant women, butonly a few studies have investigated whether music therapy relieves pregnancy-related stress and anxiety (Bauer, Victorson, Rosenbloom, Barocas, & Silver, 2010;Carolan, Barry, Gamble, Turner, & Mascareñas, 2012; Chang et al., 2008; CollinsCook, 2012; Kaufmann, 2014). The findings of Chang et al. (2008) show thatlistening to soothing music for half an hour daily significantly decreased stress,anxiety and depression after a two-week period. Bauer et al. (2010) found that livemusic and recreation therapy significantly reduced antepartum-related distress inwomen hospitalized with high-risk pregnancies. In a similar context, Nussberger(2014) found short-term live music therapy to be relaxing. Listening to music hasalso been found to increase satisfaction with nursing care: blood pressure wasreduced but did not have a significant effect on anxiety in pregnant women withpreeclampsia (Toker & Kömürcü, 2017). Sidorenko (2000) studied medical reso-nance therapy in treatment of women with high-risk pregnancies and found thata ten-day music intervention reduced the rate of premature births, reduced stressand cortisol levels, and lowered the heart rate. Moreover, the women slept betterand needed fewer painkillers, and systolic blood pressure was reduced for thosewhose blood pressure was initially high. A recently published review and meta-analysis evaluated the effectiveness of music-based interventions, and indicatedreductions in levels of stress and anxiety among pregnant women (Corbijn vanWillenswaard et al., 2017).

Music therapy may contain live or recorded music, and some studies havecompared the effects of these variations on hospitalized patients (Arnon et al.,2006; Bailey, 1983; Garunkstiene, Buinauskiene, Ingrida Ulozienec, & Markuniene,2014). According to these studies, live music therapy would be more effective andbeneficial for the patient. However, the effects of live music therapy on the HRV andself-reported stress and anxiety among hospitalized pregnant women have not beenpreviously studied.

We empirically studied the novel problem of how live music therapy affects HRVand self-reported stress and anxiety among hospitalized pregnant women. Accordingto our hypothesis, women receiving live music therapy have higher HRV than thewomen in the control group, and repetition enhances the effect of musictherapy. Second, we hypothesize that music therapy reduces the levels of self-reported stress and anxiety more than resting. Third, we hypothesize that womenwho initially are more anxious, will benefit more from the music therapy.Furthermore, we wanted to investigate whether the first and the second half of themusic therapy session might affect HRV measures differently.

Methods

Study design

The study was designed as a randomized controlled trial (RCT). On the first day ofthe experiment the pregnant women on the ward received a detailed explanation ofthe goals and procedures of the study, and thereafter gave signed informed consent.The participants were then randomly assigned to either the music therapy group(N = 52) or the control group (N = 50). The music therapist performed the allocation

NORDIC JOURNAL OF MUSIC THERAPY 9

by drawing lots. Only the music therapist saw the randomization list, and it wasstored according to the Ethical Approval. The random assignment into groups wasperformed before any data collection (e.g. questionnaires on stress and anxiety levels,or physiologic data) took place, and the group assignment was not altered during thecourse of the study. The music therapy group received live music therapy on threeconsecutive days, for half an hour at a time. The women in the control group wereinstructed by the music therapist to rest in their bed for comparable time periods andwere left alone during this time. The music therapist informed the participants in thecontrol group when the time of rest started and ended. The instruction to rest was inconjunction with the instructions given by the physician and the nurses at the ward:to spend time in the bed and try to find a comfortable resting position. Theexperienced levels of stress and anxiety were measured with Perceived Stress Scale(PSS) and State Scale of the State-Trait Anxiety Inventory (S-STAI). The participantscompleted the questionnaires both before and after the intervention. The physiolo-gical stress levels of the participants were assessed by measuring their HRV.

Participants and settings

The participants of the study (N = 102) were pregnant women (mean age 31 years, range18–47) who were hospitalized due to pregnancy-related complications. Sample size esti-mate was based on previous work in the field, reporting significant findings with similargroups of participants (Bauer et al., 2010; Chuang et al., 2012b; Toker & Kömürcü, 2017;Yang et al., 2009) and on the estimated number of participating patients on the ward. Thestudy included all the volunteers that matched the eligibility criteria: (1) a minimum age of18 years, (2) the prospect of spending four or more days in inpatient care (estimated bya nurse), and (3) fluent speaker of Finnish or Swedish. Altogether 148 participants werescreened for eligibility. Forty-six women declined to participate in the study. The mostcommon reason for declining was that the mother felt unable or unwilling to concentrateon anything new because of the stressful circumstances. Some mothers declined due tounwillingness to wear the measurement device (Figure 1).

The study took place on the antenatal ward at the Helsinki University Hospital inFinland from April 2013 until November 2014. The ward has 19 beds for antenataland labor induction patients. Patients were admitted to the ward due to variousproblems related to pregnancy. The most common diagnosis was the risk of pre-mature delivery. These patients had for instance premature rupture of membranes oruterine contractions with shortened cervix. Both pregnancies with singletons andtwins were represented. Other subgroups were women with hypertensive problemslike preeclampsia, fetal growth restriction from different origins, antenatal bleedingand maternal diabetes. The average stay on the ward was 3 days. We were not able tocollect the individual medical records of the women participating in our study, butwe estimated that there was no difference between the therapy and control groups inthe reasons for admission to the ward. Thus, the characteristics of the pregnancycomplications are unlikely to have interfered with the results.

Permission for the study

The study complies with the Declaration of Helsinki and was approved by the Ethicalcommittee of Women and Child Psychiatry in the Hospital District of Helsinki and

10 P. TECKENBERG-JANSSON ET AL.

Uusimaa (391/13/03/03/2012). The research permission was granted by the depart-ment of Women and Children Diseases of the Helsinki University Central Hospital,Hospital District of Helsinki and Uusimaa (Permission: 07.03.2013). Each participantwas informed both orally and in writing regarding the course of the study, that thecollected data would be recorded anonymously, and that their personal informationwould be kept confidential. The participants were encouraged to ask for moreinformation at any time. Written informed consent was obtained from allparticipants.

Intervention

In anthroposophical music therapy, the lyre is used as an instrument that addressesthe middle part of the human being in a way that is liberating and enhances a more

Figure 1. Flow chart of participant enrollment.

NORDIC JOURNAL OF MUSIC THERAPY 11

inward experience (Reinhold, 2012). Its gently resounding tone quality can encouragegeneral relaxation and relieve congested breathing, pain and tension. The sound ofthe lyre is gentle, soft and pleasantly enveloping. It opens up a space for sensitivelistening and creates a calm atmosphere.

This research included two different lyre instruments: a seven-stringed pentatonicAuris children´s lyre from Sweden (http://auris-musical-instruments.com/lyres) anda Tao-lyre built by R. Benedek, Germany (https://www.aeterherz.de/startseite/english/).The word tao signifies the primordial essence or fundamental nature and balance of theuniverse. Inspired by the Austrian philosopher Dr. Rudolf Steiner, the Tao-lyre wascreated as a therapeutic instrument by a music therapist Victoria Ryan in Vienna anda German constructor of instruments Andreas Lehmann. The Tao-lyre, a vibroacousticinstrument, has 48 strings but only four tones: b, a, e, d. Each tone has three strings andthe tones extend over four octaves. The sound level of these lyres is 5–68 dB. In a therapysession the sound intensity is adjusted individually to each patient for a comfortable level,and kept between 10 and 35 dB.

The live music therapy included playing of the two lyre instruments and hummingbeside the bed of the patient. The duration and intensity of the intervention wasplanned by a trained and experienced music therapist, who took into account herexperiences of the needs and preferences for music therapy in pregnant women, andthe information of the mean stay of the patients of the ward. The music was adjustedindividually to avoid any kind of overstimulation and the participants were asked ifthey had a favorite lullaby. First, the pentatonic lyre was placed and played on theabdomen of the patient who was comfortably lying on her back. Lullabies werehummed according to the patient’s wishes. The patients were also encouraged tohum along and after the first half of the therapy session they were offered theopportunity to play the lyre themselves, however, only a few were willing to do so.Thereafter, the music therapist played the Tao-lyre positioned on the legs of theparticipant. After that, the patient turned to lie on her side, and the Tao-lyre wasplayed against her back. The lyre was touching the body so that the sound wouldresonate stronger within the body. The resonance and the sound of the lyre helpedthe women to focus on their own body and the baby. This offered them anopportunity to become aware of breathing and possible tensions. Four of the parti-cipants did not wish the lyre to touch their bodies, and therefore the lyre was playedby their side.

Outcomes

HRV refers to the alteration in time between consecutive heart beats (Berntson et al.,1997). HRV was calculated from the electrocardiogram (ECG) recorded at 1000 Hzusing a single lead (two electrodes) with the Firstbeat Bodyguard device (FirstbeatTechnologies). The goal of HRV analysis is to investigate the natural, minor varia-tions in the heart rhythm. The basis for the analysis is a so-called RR (interbeat)interval time series derived from the ECG, describing the time (in milliseconds)between successive R-peaks in the ECG waveform. On the first day of the study, thedevice was attached by the music therapist to the skin of the patient: one electrodebelow the right clavicle and the other electrode on the lower left rib cage. The datawas recorded continuously during the three days of the study and the device was to

12 P. TECKENBERG-JANSSON ET AL.

remain in place, except during showering. After the three-day period, the device wasremoved.

HRV data was collected from a total of 102 participants. The full three-day data setwas available from 71 participants. Of the participants in the music therapy group 38women received the intervention during 3 days, 11 women during 2 days and 3 womenduring 1 day. In the control group the corresponding numbers were 33 women during3 days, 7 women during 2 days and 10 women during 1 day. The 31 dropouts weremostly due to the experiment ending in delivery (3 women) or the mother leaving thehospital (15 women) and in 13 cases due to skin irritation caused by the ECGelectrodes.

The participants were asked to provide personal information on a questionnairewhich consisted of nine questions concerning their age, the gestational age of thefetus, their education, profession, earlier pregnancies, number of children, reason forhospitalization, diagnosis, and the date of arrival at the hospital.

A shortened, 5-item version of The Perceived Stress Scale (PSS; Cohen, Kamarck,& Mermelstein, 1983; Pesonen et al., 2008) was used to evaluate the participants’ self-reported stress level before and after the intervention. The questions included forinstance: “In the past two weeks, how often have you felt that things were going yourway?,” “How often have you felt that you were unable to control the important thingsin your life?,” and “How often have you felt nervous and stressed?” The responses weregiven on a 5-point Likert scale (1 = never, 5 = very often).

Self-reported anxiety before and after the intervention was evaluated with TheState-Trait Anxiety Inventory (STAI, Spielberger, Gorsuch, & Lushene, 1970). Thequestionnaire is composed of 20 items each making a statement, such as “I feel calm,”“I am worried,” and “I feel confused.”Half of the statements describe positive feelings(“I feel comfortable”), and the other half negative (“I feel frightened”). The participantswere instructed to mark on a 4-point Likert scale how much they agreed with thestatement at that moment (1 = not at all, 4 = very much so).

Data preprocessing and statistical analysis

The analyses were made using R (version 3.1.1 GUI 1.65 Mavericks build; R CoreTeam, 2014). The colibri R-package (Henelius, 2014) was used to analyze the HRVdata. The sessions of music therapy (or rest in the control group) lasted 20–30 minand the HRV was analyzed from these periods. Artifacts were automatically removedusing the method by Xu and Schuckers (2001). On the basis of visual inspection ofthe data, ten measurement sessions (i.e. therapy or rest sessions) were removed due tobad signal quality.

The data from each session was divided into segments with a length of 2.5 min. Thesegment length was chosen to be shorter than the usual five minutes in order to minimizethe effects of motion artifacts due to the participant changing her position. During latepregnancy, and especially during compromised pregnancies as experienced by the parti-cipants of this study, the comfortable position of the mother is of extreme importance. Forthis reason, participants wanted to change their position often. The data segments in whichthe maximum heart rate exceeded the mean value of the data from the full measurementsession bymore than two standard deviations (SD) were excluded from the analysis as theywould likely contain excessive noise and/or artifacts. For each included segment, time-domain, frequency-domain and geometric HRVmetrics were computed. The time domain

NORDIC JOURNAL OF MUSIC THERAPY 13

metrics were: mean heart rate (i.e. beats per minute), median heart rate, RMSSD (root-mean square of successive RR-interval differences), SDNN (standard deviation of normalRR-intervals), and pNN50 (the proportion of pairs of successive normal-to-normal RR-intervals differing more than 50 ms). The frequency-domain metrics were estimated usingthe Lomb-Scargle periodogram (Scargle, 1982). The frequency domain metrics were:power in the low-frequency (LF: 0.04–0.15 Hz) and high-frequency (HF: 0.15–0.4 Hz)bands, commonly used in HRV analyses (Task Force of the European Society ofCardiology and the North American Society of Pacing and Electrophysiology, 1996). Theratio of power in the LF to HF band (LF/HF) was also computed. Additionally, thenormalized values of both LF and HF were calculated. The geometric metrics used werePoincaré plot indices, SD1 and SD2. The Poincaré plot is a visualization of the HRV signalin which each RR-interval is plotted against the preceding RR-interval. A commontechnique is to fit an ellipse to the shape of the Poincaré plot, where the dispersions (orstandard deviations) along the minor and major axis of the ellipse are referred as SD1 andSD2, respectively. SD1 is known to describe short-term and SD2 long-term variability of

Table 1. Description of different HR and HRV metrics measured during the sessions.

Metric Unit Description Reflects

Indication ofrelaxation (e.g.low mentalstress) Reference

MeanHR

bpm Mean heart rate Relative balancebetween sympatheticand parasympatheticsystems

Lower values Taelman,Vandeput,Vleminck,Spaepen, & VanHuffel, 2011

MedianHR

bpm Median heart rate

SDNN ms Standard deviation ofnormal-to-normal RR-intervals

Total HRV Higher values Taelman et al.,2011

RMSSD ms Root-mean square ofsuccessive RR-intervaldifferences

Parasympatheticactivity

Higher values Taelman et al.,2011

pNN50 % Proportion of pairs ofsuccessive normal-to-normalRR- intervals differing morethan 50 ms

Parasympatheticactivity

Higher values Taelman et al.,2011

LF ms2 Power in low-frequency band(0.04–0.15 Hz)

Both sympathetic andparasympatheticactivities

Lower values Delaney & Brodie,2000

LF.n - Normalized value of LFLF.n = LF/(LF + HF)

HF ms2 Power in high-frequencyband (0.15–0.4 Hz)

Parasympatheticactivity

Higher values Delaney & Brodie,2000

HF.n - Normalized value of HFHF.n = HF/(HF + LF)

LF/HF - The ratio of power in LF toHF band

Sympatho-vagalbalancea

Lower values Delaney & Brodie,2000

SD1 ms Poincaré plot index –Standard deviation of short-term RR-interval variability

Parasympatheticactivity

Higher values Melillo, Bracale, &Pecchia, 2011

SD2 ms Poincaré plot index –Standard deviation of long-term RR-interval variability

Both sympathetic andparasympatheticactivities

Higher values Melillo et al., 2011

a Widely accepted to describe the sympatho-vagal balance. However, this view has also been increasinglychallenged, see more, e.g. Eckberg (1997) and Billman (2013).

14 P. TECKENBERG-JANSSON ET AL.

the RR intervals. The common behavior of the different HR and HRV metrics duringrelaxation is reviewed in Table 1 (see also review article Shaffer & Ginsberg, 2017). Wheninterpreting the results, it is important to note that many of the HRV metrics stronglycorrelate with each other (Task Force of the European Society of Cardiology and the NorthAmerican Society of Pacing and Electrophysiology, 1996). Especially, in individuals with-out heart problems the calculation of SD1 and SD2 produce very similar, or identical valuesto RMSSD and SDNN respectively (Brennan, Palaniswami, & Kamen, 2001; Carrasco,Gaitán, González, & Yánez, 2001; Karmakar, Khandoker, Gubbi, & Palaniswami, 2009).Also, the frequency measures typically correlate strongly with each other [e.g., R(HF*LF) = 0.53, p < .01, Tulppo, Makikallio, Takala, Seppanen, and Huikuri (1996)] andthe non-normalized variables of HRV are more strongly correlated with HR than normal-ized values (Sacha, 2013; for a comprehensive representation of correlations between HRVmetrics measured at rest, see Tulppo et al., 1996). Therefore, to avoid increase of type IIerrors, the results are reportedwith no corrections formultiple testing (see, for instance, thereview by Sinclair, Taylor, & Hobbs, 2013).

Linear mixed models were employed for statistical comparisons using thelme4 R-package (Baayen, Davidson & Bates, 2008). The dependent variable of themodel was a HRV metric (with separate models for each HRV measure). Group(music therapy/control), measurement day (1, 2, 3), segment and age were included inthe model as independent fixed effects, and participant was included as the randomeffect. The significance of the interaction was tested using likelihood ratio tests bycomparing the model with and without the interaction. Further, the car R-package wasused to calculate the Wald chi-square test results reported below in the result section.

The within-group (therapy or control) differences in state anxiety and stressbefore and after the study were assessed using paired sample t-tests.

Next, the sample was divided into two subsamples based on the self-reported stateanxiety in the beginning of the experiment: score 40 or more representing high stateanxiety and 39 or under representing moderate state anxiety (Spielberger, Gorsuch,Lushene, Vagg, & Jacobs, 1983; Lin, Hsieh, Hsu, Fetzer, & Hsu, 2011). The high-anxiety group included 70 participants (39 in the music therapy group), and themoderate-anxiety group included 31 participants (13 in the music therapy group).

The repeated measures t-test was employed to compare the two subsamples withrespect to the effects of the music therapy on perceived stress and anxiety. Thisanalysis was also conducted for the two subsamples (two separate models).

Results

Background information

Table 2 shows the participants’ background information that was collected for thestudy. The groups did not significantly differ in terms of age, duration of thepregnancy, education, or number of children. The initial self-reported anxiety inthe music therapy group was slightly higher than in the control group. The perceivedstress estimates measured with the PSS did not differ before the intervention.Moreover, in the highly anxious subgroup, the music therapy group containedmore participants who were able to participate in only two out of three sessions.

NORDIC JOURNAL OF MUSIC THERAPY 15

Table 2. Group-level background information of participants included in the study.

Variable Music therapy Control

52 50NMean (SD) Mean (SD) t value p value

Age 30.6 (6.9) 31.0 (5.4) 0.39 .70Gestational age (in weeks) 28.7 (3.1) 28.2 (3.4) 0.89 .37Perceived stress (PSS)a – Pre 3.0 (0.8) 2.7 (0.7)c 1.76 .08Perceived stress (PSS)a – Post 2.9 (0.8) 2.7 (0.7) 0.93 .36Symptoms of anxiety (S-STAI)b – Pre 2.4 (0.6) 2.2 (0.5)c 2.20 .03*Symptoms of anxiety (S-STAI)b – Post 2.2 (0.6) 1.9 (0.4) 2.47 .02*Education N (%) N (%)

Basic education 4 (7.7) 3 (6.0)Upper secondary level 22 (42.3) 20 (40.0)Lower academic degree 12 (23.1) 12 (24.0)Higher academic degree 13 (25.0) 11 (22.0)Missing information 1 (1.9) 4 (8.0)

Number of children0 28 (53.8) 23 (46.0)1 18 (34.6) 17 (34.0)2 5 (9.6) 7 (14.0)3 1 (1.9) 3 (6.0)

a Shortened, five-item version of the Perceived Stress Scale.b The State-Trait Anxiety Inventory.c Data from 49 participants.* p < .05.

Figure 2. Changes in SD2 of the heart rate during a therapy or a control session.Note. The figure shows the mean values of the SD2 (a nonlinear, geometric parameter reflecting the deviation of the heartrate) during a therapy or a control session. The error bars illustrate the standard error of mean (SEM). Higher SD2 values aregenerally related to more relaxation (Table 1). Corresponding changes in the other 11 HRV measures are shown inSupplemental Figure 2.

16 P. TECKENBERG-JANSSON ET AL.

Heart rate variability

The linear mixed effects modeling showed that 10 of 12 HR/HRV measures did notshow significant effects. The significant effects were the following: SD2 increasedmore during the therapy sessions than during the control (rest) sessions [χ2(2) = 4.9,p = .027; Figure 2]. The HF increased more in the control condition than in thetherapy condition [χ2 = 8.31, p = .004], but the interaction was not significant withnormalized (HF.n) values [χ2 = 1.33, p = .25]. With Bonferroni adjustment the effecton SD2 is not significant.

Comparing the three sessions, 7 of the 12 HRV metrics did not show significantdifferences. The significant effects were as follows: SDNN was more decreased in latersessions of the therapy group than in the control group [χ2 = 12.3, p = .002].Moreover, the SD2 decreased from a session to another in the therapy group[χ2(2) = 11.4, p = .003]. The LF decreased between the sessions in the therapygroup more than in the control group [χ2(2) = 12.7, p = .002; Figure 3]. Thenormalized values (LF.n) also showed significant differences [χ2(2) = 7.73, p = .02].HF.n slightly increased between the sessions in the therapy group, whereas it slightlydecreased between the sessions in the control group [χ2(2) = 7.73, p = .02]. WithBonferroni adjustment the effects on LF.n is not significant. The other HRV out-comes between therapy sessions (Supplemental Table 1 and Supplemental Figure 1)and during a therapy session (Supplemental Table 2 and Supplemental Figure 2) arepresented in the supplemental information of the study.

Figure 3. Changes in LF of the heart rate over three therapy or control sessions.Note. The low frequency (LF) variation of the heart rate. Mean values (±SEM) in the three therapy or control sessions.Corresponding changes in the other 11 HRV measures are shown in Supplemental Figure 1.

NORDIC JOURNAL OF MUSIC THERAPY 17

Stress

The paired sample t-test showed that the stress did not differ significantly within thetherapy and control groups before and after the intervention: music therapy group t(1,42) = −1.31, ns. and control group t(1,35) = 0.30, ns.

Anxiety

The t-tests revealed that in the entire study sample, the state anxiety was significantlyreduced during the study period, on average by 6.2: t(1,60) = 5.016, p < 0.001 (two-tailed). Further, when the therapy and control groups were separately examined, thestate anxiety was significantly reduced in both groups: music therapy groupt(1,31) = 3.10, p < 0.004 (two-tailed) and control group t(1,28) = 4.11, p < 0.000 (two-tailed). When the amount of change in state anxiety (before-minus-after treatment)was compared between groups, there was no significant group difference [F(1,59) = 1.707, p = .73]. In the control group the mean anxiety score was −1.27before the intervention and −7.93 after the intervention, that is, the mean anxietydecreased with 6.66. The women in the therapy group had a mean score 2.09 and theanxiety decreased after the intervention (mean score −3.69), showing a reduction of5.78 units.

For the highly anxious participants, SDNN increased within a therapy session [F(1,1516) = 4.44, p = .04]. For the moderately anxious participants, mean HR decreasedwithin a therapy session [F(1,690) = 9.52, p = .02]. HRV metrics of the highly anxiousparticipants did not differ between therapy sessions. However, in the moderately anxiousgroup, meanHR, RMSSD, and pNN50 differed between the therapy and control sessions.Figure 4 shows that the second therapy and control sessions differed from the first andthe third sessions.

The level of self-reported anxiety of the highly anxious participants was reducedboth in the music therapy [t(31) = 3.74, p < .001] and control groups [t(22) = 4.09,p < .001], but there was no difference in the magnitude of the reduction between the

Figure 4. Changes in mean heart rate, RMSSD, and pNN50 in moderately anxious participants.Note. The variation of mean heart rate, RMSSD and pNN50. Mean values (±SEM) of the moderately anxious subgroup in thethree therapy or control sessions.

18 P. TECKENBERG-JANSSON ET AL.

groups [F(1,118) = 0.27, p = .60]. In contrast, the moderately anxious participantsreported no change in the levels of anxiety and stress.

Discussion

This study aimed to assess changes in stress and anxiety in hospitalized pregnantwomen both with HRV and with questionnaires with respect to receiving live musictherapy. We succeeded in collecting a valuable and unique set of data combiningrepeated physiological measurements and assessments of self-reported stress andanxiety in pregnant women. We found positive effects of live music therapy on theHRV in this group of participants. The data did not directly support the firsthypothesis: in general, HRV was not significantly greater in the group of womenwho were randomly assigned to the live music therapy group. However, out of the 12HRV measures only SD2 increased during the music therapy session, which mayindicate that live music therapy helped the pregnant hospitalized women to relaxmore during the therapy session. This is an interesting result which is in line withprevious research showing that music therapy increases HRV and can thus reduceanxiety in pregnant women (Bauer et al., 2010; Sidorenko, 2000; Yang et al., 2009). Itis important to note that the relaxation of the mother during the late part of thepregnancy, and especially in complicated pregnancies as in the case of the partici-pants in this study, is extremely important for a good birth outcome. For this reason,any method that can be safely used to help the relaxation of these mothers ismedically relevant and useful. The results also show that effects can be observedalready in a short time period of 20–30 min.

When considering the longer-term effects of live music therapy, that is, effects thatoccur from one session to the next, our hypothesis was not directly supported by thedata. LF indicated that the participants were more relaxed on successive live musictherapy sessions, which has also been found in previous studies, whereas the othermeasures did not support the hypothesis. Thus, we were not directly able to concludethe cumulative beneficial effects of participating in the music therapy on threeconsecutive days, even though one of the measures suggested a positive session-to-session effect. On the basis of this study, we cannot conclude whether repetitionenhances the effect of music therapy. Rather, the effect might be partially temporary:Even though the women become relaxed during the therapy, by the next day theymight be anxious again because of their worries and the hospital environment. Theresults comparing the three sessions might reflect differences in the baseline HRV ofeach day instead of the pure effect of music therapy, since only the data of thedifferent therapy/rest sessions from the whole intervention were included in theanalysis. In addition, the physiological responses in pregnant women may be sup-pressed compared to non-pregnant women, as some studies found no physiologicalresponse at all, although the self-reported anxiety was relieved (Chang & Chen, 2005;DiPietro, Mendelson, Williams, & Costigan, 2012). Additionally, the women withhigher gestational age may have experienced more stress (Dennis et al., 2017).

The self-reported anxiety was significantly reduced in both the music therapy andcontrol groups, but there was no difference in the amount of reduction. Moreover,when the groups were further divided by their initial level of anxiety to groups ofhigh and moderate anxiety, no significant differences were found between the musictherapy and control groups. The perceived stress was not found altered in any of the

NORDIC JOURNAL OF MUSIC THERAPY 19

comparisons. The reduction in the perceived stress may have resulted from theprogression of the pregnancy, reducing the risk of premature birth and severecomplications. It should also be noted that the condition of our participants varieshighly from one day to the next especially when the health of the fetus or the motherchanges abruptly. This may have resulted in high deviation in the self-reported stressand anxiety in our patients, and such variation may possibly be much larger than thatcaused by live music therapy.

In some previous studies, pregnant women have reported lower levels of stressand anxiety following a music therapy intervention (Chang et al., 2008), whereasother studies report no change in the anxiety levels (Toker & Kömürcü, 2017). Inthe current study, the questions concerning the self-reported stress might havebeen too general (“in the past two weeks”), and the questions about anxiety toobound to the specific moment to reflect the relaxing effect of the three shortmusic therapy sessions. The results might have been different if we hadinstructed the pregnant women to fill in the forms right after the music therapysession. In our study, they had the opportunity to fill in the questionnairesone day later.

It was assumed that the stress experienced by the participants is mostly an acutereaction to the hospitalization and the health issues of the mother and the fetus. Wedid not assess chronic stress factors, which may nevertheless have contributed to thecomplications in the pregnancy and also to the stress levels during the inpatient care(Paarlberg, Vingerhoets, Passchier, Dekker, & Van Geijn, 1995).

Study limitations

Unfortunately, despite the random assignment into the groups, the initial self-reported anxiety in the music therapy group was slightly higher than in the controlgroup, which may have affected the comparability of the groups. Moreover, theparticipants’ musical involvement or training was not asked, and musicians andnonmusicians may have different physiological responses to music (Bernardi, Porta,& Sleight, 2006). However, none of the participants reported a music-relatedprofession.

In the music therapy group the music therapist was present during the wholetherapy session whereas the participants in the control group were left alone andasked to rest in their bed. However, five of the participants in the control groupwere once found at the end of the resting time speaking on the phone, oneparticipant was watching TV, one was knitting and one had family membersvisiting.

Another limitation of the study is the dropout rate, which was not possible toavoid due to the situation of the participants. The electrodes of the HRV measure-ment device caused skin irritation for a fairly large number of women. If non-irritating electrodes could be used the women would presumably wear the devicelonger. The missing data points were taken into account in the statistical analysis.Also, breathing has been found to affect HRV (Billman, 2011), but in this study wechose not to control breathing, in order to focus on the effects of music therapy andnot on those of guided breathing.

We did not collect more specific details or diagnosis from the medical records ofthe participants. This might be one more limitation of the study. On the other hand,

20 P. TECKENBERG-JANSSON ET AL.

the amount or type of anxiety and stress are not necessarily bound to the medicalcondition of the mother or fetus. Thus, they probably have no or a minor effect onthe mother-fetus connection in this context.

Recommendations for clinical practice

The personnel on the ward appreciated the opportunity to offer patients musictherapy given by a trained music therapist as a non-pharmacological option thatmay help the women relax as well as relieve stress and anxiety. High-risk pregnantwomen, and their infants (fetuses) are at real risk of experiencing trauma, and mayface various future medical complications connected to the infant neural develop-ment, psycho-emotional development and the parent-infant attachment patterns. It isof utmost importance that pregnant women in the hospital find ways to relax.Providing music therapy to this population is aimed to prevent the high-risk preg-nancy and the premature birth from becoming a traumatic event, and to help parentsto cope better in this fragile situation. Music therapy can provide a break from thehypervigilance which is especially harmful when continuous. Music therapy couldtherefore be used in the future as common practice for pregnant women on the ward.

On the basis of this study and other previous studies, it would be a good practicefor hospital personnel to assess the stress levels of their patients on a regular basis andactively offer music therapy: live music when available, or recorded music. Thiswould ensure that the patients who are open to receive music therapy have theopportunity to do so.

Receptive music therapy does not require physical activity or much mental energyfrom the patient, so it is easy to implement even in the challenging situations thepregnant women are facing. Music therapy could also be used outside the hospitalenvironment, in outpatient care or even at home.

Suggestions for further studies

It would be important to assess the long-term effects of live music therapy, perhapswith more therapy sessions and including a longitudinal design, so that the develop-ment and health of both the mother and the child could be assessed for one year afterdelivery. It could be useful to include a baseline measurement of the HRV in theanalysis and a follow-up measurement some time after the end of the therapy. Thecontrol group could receive an alternative intervention for relaxation in order tocompare the benefits of music therapy to other relaxation methods. In the presentstudy, the empathetic presence of the music therapist may in itself have given theparticipants the experience of being nurtured, whereas the music therapist was notpresent with the resting control group. The study also prompted new researchquestions, such as whether music therapy influences the fetal HRV, and whethermusic therapy could help to relax when suffering from premature contractions, orhaving troubles with or anxiety related to breastfeeding.

Conclusions

The SD2 HRV measured from pregnant hospitalized women increased during livemusic therapy sessions, which may indicate that live music therapy helped the

NORDIC JOURNAL OF MUSIC THERAPY 21

women to relax. Relaxation is important especially for women suffering from preg-nancy complications, since relieving stress can promote the wellbeing of the motherand the fetus. Therefore, it is recommended that music therapy is offered forpregnant women.

Acknowledgments

Grant support for this study was received from Signe and Ane Gyllenberg Foundation. The funderhad no role in the study design, data collection and analysis, decision to publish, or preparation ofthe manuscript.

Disclosure statement

The authors of the study declare that they have no conflict of interest regarding the study.

Notes on contributors

Pia Teckenberg-Jansson, music therapist, got her music therapy education and exam at theMusiktherapeutische Arbeitsstätte in Berlin, Germany, in 2004. She is working in Helsinki on anantenatal ward with pregnant women, in various nursing homes for elderly suffering from dementiaand also in palliative care. Previous research projects of music therapy for prematurely bornchildren.

Siiri Turunen received her Master of Psychology degree at University of Helsinki, Finland, in 2016,and her M.Phil. degree in cognitive science in 2013. She is currently working as a clinical psychol-ogist at a psychiatric hospital. She has also worked among children and families.

Tarja Pölkki, R.N., Ph.D., Adjunct Professor, is working as a specialist in clinical nursing science inthe unit of Children and Women in the Oulu University Hospital, Finland. Her research interestsinclude pain assessment and pain management, focused especially on non-pharmacological meth-ods, among pediatric patients, as well as family-centered care. Tarja is also a member of PEARL(Pain in Early Life) research group.

Minna-Johanna Lauri-Haikala received her Master of Psychology degree from University ofHelsinki, Finland, in 2016. She is currently working as a school psychologist. She has also workedat a children’s hospital.

Jari Lipsanen received his Master of Psychology degree from University of Helsinki, Finland, in2006. Currently working as lecturer of psychometrics in Department of Psychology and Logopedics,in University of Helsinki.

Andreas Henelius received his D.Sc. (Tech.) degree in computer science from Aalto University in2017. His current research interests include the use of randomization methods in data mining.

Ansa Aitokallio-Tallberg M.D., Ph.D. Specialist in Obstetrics, Gynecology and PerinatologyWorking as senior consultant on department of Obstetrics and Gynecology at University hospital inHelsinki.

Satu Pakarinen received her Doctor of Psychology degree from University of Helsinki, Finland, in2011. Her current research focuses on measuring the effects of acute and prolonged stress on humanphysiology and performance as well as on health and wellbeing.

Marianne Leinikka received her M.Sc. (Tech.) degree in bioinformation technology from AaltoUniversity, School of Electrical Engineering, in 2014. Her expertise is in the experimental work inphysiological measurements both in laboratory as well as real-life settings and in the analysis ofthem.

22 P. TECKENBERG-JANSSON ET AL.

Minna Huotilainen is a Professor of Educational Sciences at University of Helsinki, Finland. She isbest known for her work in studying auditory brain responses in infants and fetuses and showing thebenefits of musical training for children and adolescents as well as for her studies on neuroscience ofwork.

References

Arnon, S., Shapsa, A., Forman, L., Regey, R., Bauer, S., Litmanovitz, I., & Dolfin, T. (2006). Live musicis beneficial to preterm infants in the neonatal intensive care unit environment. Birth, 33(2),131–136.

Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008). Mixed-effects modeling with crossed randomeffects for subjects and items. Journal of Memory and Language, 59(4), 390–412.

Bailey, L. M. (1983). The effects of live music versus tape-recorded music on hospitalized cancerpatients. Music Therapy, 3(1), 17–28.

Bastani, F., Hidarnia, A., Kazemnejad, A., Vafaei, M., & Kashanian, M. (2005). A randomized controlledtrial of the effects of applied relaxation training on reducing anxiety and perceived stress in pregnantwomen. Journal of Midwifery & Women’s Health, 50(4), e36–e40. doi:10.1016/j.jmwh.2004.11.008.

Bauer, C., Victorson, D., Rosenbloom, S., Barocas, J., & Silver, R. (2010). Alleviating distress duringantepartum hospitalisation: A randomised, controlled trial of music and recreation therapy.Journal of Women’s Health, 19(3), 523–530.

Benedek, R. The tao-instruments from Benedek. Retrieved from https://www.aeterherz.de/startseite/english/

Bernardi, L., Porta, C., & Sleight, P. (2006). Cardiovascular, cerebrovascular, and respiratory changesinduced by different types of music in musicians and non-musicians: The importance of silence.Heart (British Cardiac Society), 92(4), 445–452.

Berntson, G. G., Thomas Bigger, J., Eckberg, D. L., Grossman, P., Kaufmann, P. G., Malik, M., …van der Molen, M. W. (1997). Heart rate variability: Origins, methods, and interpretive caveats.Psychophysiology, 34(6), 623–648.

Bieligmeyer, S. (2015). Klangraum Körper – Einfluss einer musiktherapeutischen Intervention mit derKlangliege auf Resilienz und Wohlbefinden in der Onkologie [Effects of a sound-bed-intervention onpsychological well-being and resilience in oncology](Unpublished doctoral dissertation). Universityof Tübingen, Germany.

Billman, G. E. (2011). Heart rate variability – a historical perspective. Frontiers in Physiology, 2, 86.Billman, G. E. (2013). The LF/HF ratio does not accurately measure cardiac sympatho-vagal balance.

Frontiers in Physiology, 4. doi:10.3389/fphys.2013.00026.Brennan, M., Palaniswami, M., & Kamen, P. (2001). Do existing measures of Poincare plot geometry

reflect nonlinear features of heart rate variability? IEEE Transactions on Biomedical Engineering,48(11), 1342–1347.

Burrai, F., Hasan, W., Fancourt, D., Luppi, M., & DiSomma, S. (2016). A randomized controlled trialof listening to recorded music for heart failure patients: Study protocol. Holistic Nursing Practice,30(2), 102–115.

Calderon-Margalit, R., Qiu, C., Ornoy, A., Siscovick, D. S., & Williams, M. A. (2009). Risk ofpreterm delivery and other adverse perinatal outcomes in relation to maternal use of psycho-tropic medications during pregnancy. American Journal of Obstetrics and Gynecology, 201(6),579.e1–8. doi:10.1016/j.ajog.2009.06.061.

Carolan, M., Barry, M., Gamble, M., Turner, K., & Mascareñas, O. (2012). The limerick lullabyproject: An intervention to relieve prenatal stress. Midwifery, 28(2), 173–180.

Carrasco, M. J., Gaitán, R., González, O., & Yánez, S. (2001). Correlation among Poincare plotindexes and time and frequency domain measures of heart rate variability. Journal of MedicalEngineering & Technology, 25(6), 240–248.

Chang, H., Yu, C., Chen, S., & Chen, C. (2015). The effects of music listening on psychosocial stressand maternal-fetal attachment during pregnancy. Complementary Therapies in Medicine, 23(4),509–515.

Chang, M., Chen, C., & Huang, K. (2008). Effects of music therapy on psychological health ofwomen during pregnancy. Journal of Clinical Nursing, 17(19), 2580–2587.

NORDIC JOURNAL OF MUSIC THERAPY 23

Chang, S.-C., & Chen, C.-H. (2005). Effects of music therapy on women’s physiologic measures,anxiety, and satisfaction during cesarean delivery. Research in Nursing & Health, 28(6),453–461.

Chuang, L.-L., Lin, L.-C., Cheng, P.-J., Chen, C.-H., Wu, S.-C., & Chang, C.-L. (2012a). Effects ofa relaxation training programme on immediate and prolonged stress responses in women withpreterm labour. Journal of Advanced Nursing, 68(1), 170–180. doi:10.1111/j.1365-2648.2011.05765.x.

Chuang, L.-L., Lin, L.-C., Cheng, P.-J., Chen, C.-H., Wu, S.-C., & Chang, C.-L. (2012b). Theeffectiveness of a relaxation training program for women with preterm labour on pregnancyoutcomes: A controlled clinical trial. International Journal of Nursing Studies, 49(3), 257–264.doi:10.1016/j.ijnurstu.2011.09.007.

Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal ofHealth and Social Behavior, 24(4), 385–396.

Collins Cook, E. (2012). The uses of music and music therapy to decrease stress and anxiety duringpregnancy (Unpublished master’s thesis). Faculty of the College of Nursing and HealthProfessions Drexel University, Philadelphia, PA.

Corbijn van Willenswaard, K., Lynn, F., McNeill, J., McQueen, K., Dennis, C.-L., Lobel, M., &Alderdice, F. (2017). Music interventions to reduce stress and anxiety in pregnancy: A systematicreview and meta-analysis. BMC Psychiatry, 17, 271. doi:10.1186/s12888-017-1432-x.

Delaney, J. P. A., & Brodie, D. A. (2000). Effects of short-term psychological stress on the time andfrequency domains of heart-rate variability. Perceptual and Motor Skills, 91(2), 515–524.

Dennis, C. L., Falah-Hassani, K., & Shiri, R. (2017, May). Prevalence of antenatal and postnatalanxiety: Systematic review and meta-analysis. The British Journal of Psychiatry, 210(5), 315–323.doi:10.1192/bjp.bp.116.187179.

Dipietro, J. A., Costigan, K. A., Nelson, P., Gurewitsch, E. D., & Laudenslager, M. L. (2008). Fetalresponses to induced maternal relaxation during pregnancy. Biological Psychology, 77, 11–19.

DiPietro, J. A., Hodgson, D. M., Costigan, K. A., Hilton, S. C., & Johnson, T. R. (1996). Fetalneurobehavioral development. Child Development, 67(5), 2553–2567. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9022256.

DiPietro, J. A., Mendelson, T., Williams, E. L., & Costigan, K. A. (2012). Physiological bluntingduring pregnancy extends to induced relaxation. Biological Psychology, 89(1), 14–20.

Dole, N., Savitz, D. A., Hertz-Picciotto, I., Siega-Riz, A. M., McMahon, M. J., & Buekens, P. (2003).Maternal stress and preterm birth. American Journal of Epidemiology, 157(1), 14–24.

Douglas, A. J. (2010). Baby on board: Do responses to stress in the maternal brain mediate adversepregnancy outcome? Frontiers in Neuroendocrinology, 31(3), 359–376. doi:10.1016/j.yfrne.2010.05.002.

Eckberg, D. L. (1997). Sympathovagal balance. Circulation, 96(9), 3224–3232.Garunkstiene, R., Buinauskiene, J., Ingrida Ulozienec, I., & Markuniene, E. (2014). Controlled trial

of live versus recorded lullabies in preterm infants. Nordic Journal of Music Therapy, 23(1), 71–88.doi:10.1080/08098131.2013.809783.

Geller, P. A. (2004). Pregnancy as a stressful life event. CNS Spectrums, 9(3), 188–197. Retrievedfrom http://www.ncbi.nlm.nih.gov/pubmed/14999159.

Gilboa, A. (2014). The dual nature of the womb and its implications for music therapy.Nordic Journal ofMusic Therapy, 23(3), 242–262. doi:10.1080/08098131.2013.809784.

Graignic-Philippe, R., & Tordjman, S. (2009). Effets du stress pendant la grossesse sur ledéveloppement du bébé et de l’enfant [Effects of stress during pregnancy on infant and childdevelopment]. Archives De Pédiatrie, 16(10), 1355–1363.

Henelius, A. (2014). Colibri: Platform for automating analyses of physiological time series in R.R-package version 0.1.2. Retrieved from https://github.com/bwrc/colibri.

Karmakar, C. K., Khandoker, A. H., Gubbi, J., & Palaniswami, M. (2009). Complex correlationmeasure: A novel descriptor for Poincaré plot. Biomedical Engineering Online, 8(1), 17.

Kaufmann, J. (2014). Musiktherapie mit risikoschwangeren Frauen [Music therapy treatmentoptions for women with high-risk pregnancies]. In J. Kaufmann, R. Nussberger, M. Esslinger,& M. M. Leitgeb (Eds.), Gespürt – gehört – gebor(g)en. Musiktherapie mit risikoschwangerenFrauen, Säuglingen und Kleinkindern [Felt – heard – born(e). Music therapy for women with high-risk pregnancies, infants and small children] (pp. 9–68). Wiesbaden: Reichert.

Kemper, K. J., & Danhauer, S. C. (2005). Music as therapy. Southern Medical Journal, 98(3), 282–288.

24 P. TECKENBERG-JANSSON ET AL.

Lai, H. L., & Li, Y.M. (2011). The effect ofmusic on biochemical markers and self perceived stress amongfirst-line nurses: A randomized controlled crossover trial. Journal of Advanced Nursing, 67(11),2414–2424.

Lin, M. F., Hsieh, Y., Hsu, Y., Fetzer, S., & Hsu, M. (2011). A randomised controlled trial of theeffect of music therapy and verbal relaxation on chemotherapy-induced anxiety. Journal ofClinical Nursing, 20(7–8), 988–989.

Lobel, M., Cannella, D. L., Graham, J. E., DeVincent, C., Schneider, J., &Meyer, B. A. (2008). Pregnancy-specific stress, prenatal health behaviors, and birth outcomes. Health Psychology, 27(5), 604–615.

Maloni, J. A., & Park, S. (2005). Postpartum symptoms after antepartum bed rest. Journal ofObstetric, Gynecologic & Neonatal Nursing, 34(2), 163–171.

Melillo, P., Bracale, M., & Pecchia, L. (2011). Nonlinear heart rate variability features for real-lifestress detection. Case study: Students under stress due to university examination. BiomedicalEngineering Online, 10(1), 96.

Mulder, E. J., Robles de Medina, P., Huizink, A., Van Den Bergh, B. R. H., Buitelaar, J., &Visser, G. H. (2002). Prenatal maternal stress: Effects on pregnancy and the (unborn) child.Early Human Development, 70(1–2), 3–14. doi:10.1016/S0378-3782(02)00075-0.

Nussberger, R. (2014). Impuls-Musiktherapie mit hospitalisierten risikoschwangeren Frauen undihren ungeborenen Kindern [Short-term music therapy involving hospitalised women with high-risk pregnancies and their unborn children]. In J. Kaufmann, R. Nussberger, M. Esslinger, &M. M. Leitgeb (Eds.), Gespürt – gehört – gebor(g)en. Musiktherapie mit risikoschwangeren Frauen,Säuglingen und Kleinkindern [Felt – hear – born(e). Music therapy for women with high-riskpregnancies, infants and small children] (pp. 69–128). Wiesbaden: Reichert.

Paarlberg, K. M., Vingerhoets, A. J. J. M., Passchier, J., Dekker, G. A., & Van Geijn, H. P. (1995).Psychosocial factors and pregnancy outcome: A review with emphasis on methodological issues.Journal of Psychosomatic Research, 39(5), 563–595.

Pesonen, A., Räikkonen, K., Heinonen, K., Komsi, N., Järvenpää, A., & Strandberg, T. (2008).A transactional model of temperamental development: Evidence of a relationship between childtemperament and maternal stress over five years. Social Development, 17(2), 326–340.

R Development Core Team (2014). R: A language and environment for statistical computing. RFoundation for Statistical Computing, Vienna, Austria. Retrieved from https://www.R-project.org/.

Räikkönen, K., Pesonen, A.-K., Heinonen, K., Kajantie, E., Hovi, P., Järvenpää, A.-L., …Andersson, S. (2008). Depression in young adults with very low birth weight: The Helsinkistudy of very low-birth-weight adults. Archives of General Psychiatry, 65(3), 290–296.doi:10.1001/archgenpsychiatry.2007.40.

Reinhold, S. (2012). Anthroposophical Music therapy. Filderstadt: Verlag Freie Ausbildung.Sacha, J. (2013). Why should one normalize heart rate variability with respect to average heart rate.

Frontiers in Physiology, 4. doi:10.3389/fphys.2013.00306.Scargle, J. (1982). Studies in astronomical time series analysis. II-Statistical aspects of spectral

analysis of unevenly spaced data. The Astrophysical Journal, 263, 835–853.Shaffer, F., & Ginsberg, J. P. (2017). An overview of heart rate variability metrics and norms.

Frontiers in Public Health, 5, 258.Sidorenko, V. N. (2000). Clinical application of medical resonance therapy music in high-risk

pregnancies. Integrative Physiolocigal and Behavioral Science, 35(3), 199–207.Sinclair, J., Taylor, P. J., & Hobbs, S. J. (2013). Alpha level adjustments for multiple dependent

variable analyses and their applicability – A review. International Journal of Sports Science andEngineering, 7(1), 17–20.

Soleimani, F., Zaheri, F., & Abdi, F. (2014, Jun). Long-term neurodevelopmental outcomes afterpreterm birth. Iran Red Crescent Medical Journal, 16(6), e17965. doi:10.5812/ircmj.17965.

Spielberger, C., Gorsuch, R., & Lushene, R. (1970). STAI manual for the state-trait anxiety inventory.Palo Alto, CA: Consulting Psychologists.

Taelman, J., Vandeput, S., Vleminck, E., Spaepen, A., & Van Huffel, S. (2011). Instantaneous changesin heart rate regulation due to mental load in simulated office work. European Journal of AppliedPhysiology, 111(7), 1497–1505.

Task Force of the European Society of Cardiology and the North American Society of Pacing andElectrophysiology. (1996). Heart rate variability: Standards of measurement, physiological inter-pretation and clinical use. Circulation, 93 (1043–1065). doi:10.1161/01.CIR.93.5.1043.

NORDIC JOURNAL OF MUSIC THERAPY 25

Toker, E., & Kömürcü, N. (2017). Effect of Turkish classical music on prenatal anxiety andsatisfaction: A randomized controlled trial in pregnant women with pre-eclampsia.Complementary Therapies in Medicine, 30, 1–9.

Tulppo, M. P., Makikallio, T. H., Takala, T. E., Seppanen, T. H. H. V., & Huikuri, H. V. (1996).Quantitative beat-to-beat analysis of heart rate dynamics during exercise. American Journal ofPhysiology-Heart and Circulatory Physiology, 271(1), H244–H252.

Van den Bergh, B., Mulder, E., Mennes, M., & Glover, V. (2005). Antenatal maternal anxiety andstress and the neurobehavioural development of the fetus and child: Links and possible mechan-isms. A review. Neuroscience and Biobehavioral Reviews, 29(2), 237–258.

Xu, X., & Schuckers, S. (2001). Automatic detection of artifacts in heart period data. Journal ofElectrocardiology, 34(4), 205–210.

Yang, M., Li, L., Zhu, H., Alexander, I. M., Liu, S., Zhou, W., & Ren, X. (2009). Music therapy torelieve anxiety in pregnant women on bedrest: A randomized, controlled trial. MCN: TheAmerican Journal of Maternal/Child Nursing, 34(5), 316–323.

26 P. TECKENBERG-JANSSON ET AL.